Sie sind auf Seite 1von 13

Applied Ethics in Internet Research

Edited by

May Thorseth

Q PfO^p^*^ 2
ftr
Applied QVitcs cu^uj

/pnnW »'
/vj
AjortAJcty by

Programme for Applied Ethics


Publication Series No. 1
Norwegian University of Science and Technology, Trondheim
Generating data online: ethical concerns and challenges for the C21 researcher

Chris Mann

Generating data online: ethical concerns and challenges for the C21
researcher

Within the social sciences codes of ethics have developed over time. Gradu-
ally informal arrangements between researcher and participants about confi-
dentiality of data and respect for privacy have become highly formalised (see
discussion in Adler and Adler, 2001). There are now various kinds of ethical
'rule-books' such as discipline based professional codes and funding body re-
quirements. A very influential example of organisations set up to give ethical
guidelines is the Institutional Review Boards (IRBs) in the USA. Following
publicity about some notorious research studies, such as, the Milgram experi-
ment in 1965 (ibid. p. 517), the American government decided that individual
researchers would no longer decide for themselves about what would be an
ethical way to act in a given situation. Instead there would be public discus-
sions about the rights and wrongs of individual research studies. Institutional
Review Boards, which have become very common since 1990, are concerned
w i t h the moral good of the country, the power of law, and the legal protection
of universities and other research centres. To meet their mandate, IRBs and
other forms of ethical review boards need to work with well-defined 'duty' or
rule driven ethics. For many, the over-riding duty is the directive not to harm
others. Indeed, not harming others is the core value in social science research
and has been incorporated into guidelines for ethical practice. These guide-
lines include such rules as not invading privacy, briefing participants about the
nature of the research and protecting them from risks associated with the re-
search.
The question raised by Internet research is whether such well established
ethics codes are appropriate for work online. Some commentators feel that
there is a tension between traditional guidelines for research and the new tech-
nologies.

31
Chris Mann

Should we really expect that the normative vocabulary that has served us
in more traditional communities - including scientific communities - w i l l au-
tomatically serve us equally well in their electronic counterparts? (Jones,
1994)
This question is important because Internet research has certain distinc-
tive features that may increase the challenge of acting ethically - features such
as:
• the blurred distinction between public and private domains
• the ease of anonymity or pseudonymity
• the geographical dispersal of users
• the ability to record and archive communications without consent
• the ability to track participants using the technology
In the following six sections we shall raise the ethical problems and debates
associated with these features and suggest possible ways forward for research-
ing ethically online.
1. Ethical issues linked to the focus of research on the Internet
2. Ethical issues linked to using the Internet as a tool to generate data
3. Key ethical debates
4. Assessment of risk in Internet research
5. The way forward?
6. Conclusions
We w i l l begin by looking at possible ethical issues that are raised by choosing
to focus research on aspects o f the Internet and then we w i l l turn to ethical is-
sues associated with using the Internet as a tool for research regardless o f the
research focus.

1 Ethical issues linked to the focus of research on the Internet

Researching Computer Mediated Communication (CMC)


A large body of literature focuses on Computer Mediated Communication
(CMC) (see Mann and Stewart, 2000). C M C refers to the ways we communi-
cate using Internet technology - using email or chat rooms for instance. CMC
allows us to investigate human communication when there are limitations - or
perhaps opportunities? - in terms of having no visual clues about the people
we are communicating with. In this area research questions might ask: How
do people communicate when they can't see or hear each other? How do rela-
tionships flourish when immediate evidence of gender or age or ethnicity is
not getting in the way? How does the way something is written - or how it is
read - trigger anger or affection or agreement/disagreement between people?
The Internet may also be a good location for forms o f interaction that are diffi-
I
Generating data online: ethical concerns and challenges for the C21 researcher

that has served us cult to get access to in the embodied world. For instance, the Internet proved a
lunities - w i l l au- good site for studying the transmission of rumours - because they are passed
lterparts? (Jones, round newsgroups and chatrooms in a text form that is easily accessed. On the
face of it analysis, such as discourse analysis, of the written text produced by
is certain distinc- C M C is fairly innocuous research in terms of ethics. But the research ques-
tly - features such tions above remind us that communication between humans links directly to
relationships and hence our deepest feelings - suggesting that, in some cir-
mains cumstances, ethical issues w i l l arise (see debates in section 3).

Studies of virtual communities


hout consent A second area of research focusing on the Internet is the study of online
worlds, both fantasy worlds and also worlds of shared experience or interests
lems and debates (see Mann and Stewart, 2000). Life online gives us an unprecedented opportu-
vard for research- nity to explore the ways humans create and maintain culture. Researchers
have asked if relationships in the virtual world, away from many of the rules
nternet and norms o f the embodied world, w i l l be different. Online we are not just
generate data without a teacher or parent - or policeman! - to insist on rules - we are also
without our bodies and, if we choose it, we are without our histories. Some
online communities have already developed their own norms and laws that
may, or may not, echo those of the embodied world. But there are ethical is-
sues here. If some online venues A R E 'separate' worlds which can apparently
aised by choosing reject many real life moral norms - is it a case o f 'anything goes' for the re-
turn to ethical is- searcher too? If researchers as ethnographers sink themselves into the virtual
t regardless of the 'culture' may they feel free to lose moral scruples - perhaps raiding the virtual
world to plunder data if they can get away w i t h it? A n d how should a re-
searcher act in an online world where there A R E newly formed rules? Among
other things chat rooms or newsgroups might require incomers to declare
themselves and their business. Or they might share a tacit understanding - a
Internet
group culture - that sets a value on openness and mutual trust. Does the re-
searcher as moral agent adopt the rules of that online world? For instance, if a
certain newsgroup exists so that people can open up and share their thoughts
i Communication
and feeling about a particular issue - is the researcher morally obliged to do so
ays we communi-
too? Researchers conducting C M C might take different positions in the three
br instance. C M C
debates discussed in section 3 depending on their views about this.
ire limitations - or
about the people
s might ask: H o w The impact of the Internet use on everyday life
iier? H o w do rela- A wide range of disciplines such as sociology, psychology, law, politics and
ige or ethnicity is health research have looked at the impact of Internet use on everyday life. Just
itten - or how it is to give an idea of the scope of work that has been done here is an overview of
: between people? work that has been of interest to psychologists:
;tion that are diffi-

33
Chris Mann Generating

• group dynamics online - conformity, polarization, conflict, co-opera- ward here is to use
tion ered about people
• experimenting with online personae in MUDS and MOOS itself.
• psychology of aggression online Researchers
• interpersonal relationships online - friendships/romance that does not repre
• psychological impact of online pornography etc no access to the Ir
• personal 'impression management' - personal Web pages etc quiry would be mo
• impact of Internet use on state of health - Internet addiction information poor -
• support/altruism on the Net - especially with stigmatised groups two.
• gender roles, stereotypes, conflicts on the Internet
Experiments
Research focusing on the impact of the Internet may be conducted using tradi- At this point in tin
tional methods - or using online methods. Both qualitative and quantitative re- good science - and
search methods in conventional research have a history of ethical challenges - life laboratory work
and ethical precepts - that have been developed over time (see Mann and (Binbaum, 2000).
Stewart, 2000). We shall now turn to ethical issues that are raised once we like schematic imag
move these traditional research methods online. laboratory work eve
answers when respo
'responses were due
2 Ethical issues linked to using the Internet as a tool to generate effect of experimenta
It has also been
data
ethical than face-to
Researchers may use the Internet as a tool to generate data on a wide range of obvious source of c<
issues in many subject areas. In practical terms the Internet can be used di- the researcher. The
rectly - by asking participants to respond electronically to research queries. It an experiment even
can also be used indirectly by gathering ongoing or archival messages or rewarding. So to lose
'posts' arising from chat rooms or newsgroups. 'nontrivial benefit of

Survey approaches
Qualitative interview
Given that only approximately 0.01% of the world's population were online at The exact opposite
the start of 2000, the patchiness of Internet access has clear implications for into play when consi
research approaches that seek representative sampling. In statistical terms ing. In qualitative n
large samples (which are certainly possible with e-mail and, in particular, adaptable, flexible ii
web-page-based surveys) do not mean anything unless they are representative and understanding' (
of a target population. The skewed nature of Internet use worldwide means cellent data by beinj
that general mailings will probably not be representative. This bias in sam- and able to engage
pling gives statistical work low 'power' and can undermine the ethics of a cannot achieve this 1
study by providing inadequate or incorrect data (see Mann and Stewart, 2001). nication - and that th
For instance, early tests of logical reasoning conducted online were probably The jury is still out
affected by the fact that at that time a large percentage of Internet users were committed to the re
computer scientists with extensive training in symbolic logic! There are ways equal to that product
round this - for instance researchers can target specific groups. One way for-

34
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

• group dynamics online - conformity, polarization, conflict, co-opera- ward here is to use databases of names and addresses which have been gath-
tion ered about people without their consent - but this is an ethical quandary in
• experimenting with online personae in MUDS and MOOS itself.
• psychology of aggression online Researchers may also ask whether is it ethical to use a research method
• interpersonal relationships online - friendships/romance that does not represent the many groups of people and parts of the world with
• psychological impact of online pornography etc no access to the Internet. It could be argued, for instance, that scientific in-
• personal 'impression management' - personal Web pages etc quiry would be more focused on the interests of the information rich than the
• impact of Internet use on state of health - Internet addiction information poor - which could reinforce existing social divisions between the
• support/altruism on the Net - especially with stigmatised groups two.
• gender roles, stereotypes, conflicts on the Internet
Experiments
Research focusing on the impact of the Internet may be conducted using tradi- At this point in time experimental work conducted online is only seen to be
tional methods - or using online methods. Both qualitative and quantitative re- good science - and hence ethical - if it is comparable in terms of results to real
search methods in conventional research have a history of ethical challenges - life laboratory work. Results in this area are at an early stage but seem positive
and ethical precepts - that have been developed over time (see Mann and (Binbaum, 2000). Some research presenting online participants with cartoon-
Stewart, 2000). We shall now turn to ethical issues that are raised once we like schematic images of human bodies or faces produced similar reponses to
move these traditional research methods online. laboratory work even working cross-culturally. In one study the closeness in
answers when responding to 'the attractiveness of the female form' suggested
'responses were due to the same psychological variables and that there was no
2 Ethical issues linked to using the Internet as a tool to generate effect of experimental medium' (Senior and Smith, 1999, p.443).
It has also been suggested that appropriate online studies may be more
data
ethical than face-to face experiments because the Internet removes 'the most
Researchers may use the Internet as a tool to generate data on a wide range of obvious source of coercion' in the experiment - that is the physical presence of
issues in many subject areas. In practical terms the Internet can be used di- the researcher. The researcher's presence may lead someone to continue with
rectly - by asking participants to respond electronically to research queries. It an experiment even though they feel uncomfortable or find the experiment un-
can also be used indirectly by gathering ongoing or archival messages or rewarding. So to lose this negative aspect of face-to-face work is seen to be a
'posts' arising from chat rooms or newsgroups. 'nontrivial benefit of Internet-based research' (Nosek et al, 2002, p. 162).

Survey approaches Qualitative interviews


Given that only approximately 0.01% of the world's population were online at The exact opposite concern regarding the presence of the researcher comes
the start of 2000, the patchiness of Internet access has clear implications for into play when considering the 'good science' aspects of qualitative interview-
research approaches that seek representative sampling. In statistical terms ing. In qualitative research the face-to-face interviewer is, ideally a 'smart,
large samples (which are certainly possible with e-mail and, in particular, adaptable, flexible instrument who can respond to situations with skill, tact,
web-page-based surveys) do not mean anything unless they are representative and understanding' (Seidman, 1991: 16). Such an interviewer would evoke ex-
of a target population. The skewed nature of Internet use worldwide means cellent data by being a careful listener, non-judgemental, perceptive, focused
that general mailings will probably not be representative. This bias in sam- and able to engage with interviewee(s). The fear is that online interviewing
pling gives statistical work low 'power' and can undermine the ethics of a cannot achieve this level of highly interactive, rich and spontaneous commu-
study by providing inadequate or incorrect data (see Mann and Stewart, 2001). nication - and that the data produced will be inferior and so ethically unsound.
For instance, early tests of logical reasoning conducted online were probably The jury is still out about this - but many believe that once participants are
affected by the fact that at that time a large percentage of Internet users were committed to the research, online interviewers can gather data which is the
computer scientists with extensive training in symbolic logic! There are ways equal to that produced face to face (see Mann and Stewart, 2000 and 2001).
round this - for instance researchers can target specific groups. One way for-

34 35
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

However, the possibly 'coercive' aspects of the research process - the im- 3 Key ethical debates
pact of power relations between people - is also a feature of all kinds of inter-
viewing techniques. Working face-to-face, status, age, gender and ethnicity (a) The public-private debate
limit the sort of relationship that can be developed with participants and may A key issue in Internet research is whether the online world is a 'public' or 'pri-
also restrict access to research settings and to data. Online, these limits need vate' place - or in more sophisticated formulations 'private places in a public
not apply and excellent data may be produced as a result. However, although forum'. One definition of 'private' is that an individual can reasonably expect
the potential for reducing power differentials seems to be a bonus for Internet that no observation or recording is taking place. Is cyberspace a context within
research - the possibility of manipulating power relations online is an ethical which anyone can "reasonably expect" that no researcher is observing or re-
minefield. The Internet is a medium which allows for deception - and re- cording what you say? Certainly the naive Internet user is inclined to think
searchers face ethical decisions regarding the amount of information about that messages they circulate are only read by people like themselves. Whereas
themselves they should make known. The psychiatrist who masqueraded as a in fact, as we know, they can be read by all kinds of people.
woman to tap intimate confessional stories is a case in point. His persona, "Don't tell my boss about my cocaine habit" is typical of this kind of In-
known as Joan, was adopted when he discovered that women were much more ternet posting, shared only with ten million of one's closest friends (Covering-
open and intimate with him when he changed the power differential between ton, 1995, p.33).
them. As a man - and with status as a psychiatrist - he may have gathered less But even when individuals are not being particularly indiscrete they may
data than he did in his computer identity as a handicapped woman. But when still be taken aback to find that a researcher is not only reading their messages
his true identity was revealed the women he had talked to were aghast and saw but also keeping copies of what they say.
him as a cheat and a fraud (Van Gelder, 1991). Let us look at differences between research online and research in situa-
tions where people are in 'public' situations in real life. In real life naturalistic
Unobtrusive observation observation in public settings usually does not require consent. Why is this?
Within traditional social science research the general rule is that behaviour in Partly it is the sheer difficulty of getting consent from, for instance, a football
public places is a legitimate object of scrutiny for the social researcher, crowd or people in a church. Partly because it is assumed that if someone is
whereas that in private is not, unless consent is given. However, ethical com- doing something in public they don't mind other people seeing them doing it.
mittees may allow unobtrusive observation of behaviour performed in the And even if someone is 'caught out' doing something silly in public by a
public domain to go ahead without consent if there is no risk to participants - nameless person in the crowd who doesn't know them and can't follow it up -
and if the researcher can show that it is essential for the interaction under scru- well it's hard luck but it doesn't matter that much. Thus the freedom given by
tiny to remain 'natural' and not contaminated by the researcher's agenda. This not being known in a public place is important.
research approach, variously called naturalistic studies or participant observa- Real life certainly has 'private spaces in public places'. There are physical
tion, is seen to move easily to the Internet (see Mann and Stewart, 200o). It clues to this - the way seats are grouped in a hotel lounge, the plastic hoods
can open up new possibilities because researchers can infiltrate newsgroups sometimes found over public phones - all give tacit privacy. There are also
and chatrooms and 'watch' what is going on while pretending to be one of the customs about privacy. If you are having a picnic on a beach you would not
group - it is temptingly easy to gather information without people having any expect someone to join you without invitation - or even to walk across the
idea that they are part of a research project. In real life it may be impossible to patch of sand that is your 'tablecloth' People in public life might see you and
carry off membership of some groups (even your haircut might forbid it) but hear you but the world goes on much of the time as if they did not. If a couple
working online researchers can enter many kinds of groups without difficulty. are having an argument at the next table to you in a restaurant they rarely wel-
However, while there is no doubt that as a practical method unobserved ob- come you leaning over to give advice.
servation is ideally suited to Internet research - and may garner excellent data So let us consider research in a public place. In the real world someone
- there are ethical drawbacks. In fact it is in this area above all that three of the with a notebook or a good memory can be watching what's going on in a pub-
key ethical debates about Internet research arise; these are public-private is- lic area (can even be part of what's happening) and can go home and write an
sues, the ownership of one's own words, and concepts of risk. We shall now interesting paper on, for instance, mating rituals in a bar or mobile phone rage
look at these debates in more detail. in a theatre. But let us take this further. How would you feel about this re-

36 37
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

search activity if the researcher was tapeing the argument in the restaurant or • demand little in the way of individual commitment or identification
videoing the flirting in the bar? What is different here? Why does this feel • are often open to anyone
much more like an invasion of privacy? These are important questions be- • demand no rite of passage
cause in virtual worlds everything which constitutes the world - that is the text • have little in the way of common tradition or heritage
through which virtual worlds exist - can be recorded by the researcher. Ob- • make it possible to link data sources in a way that reveals far more
serving virtual worlds is not an impressionistic method - it lays bare every- about a subject than what he or she might "voluntarily consent" to.
thing that happens. In virtual worlds there are often references to town halls, He concludes by saying that it is these features that make such groups fasc
villages and cafes - communal spaces - suggesting that the interaction which nating subjects to study - and that the application of rule based ethics with n
occurs within them is also public - and thus falls within the remit of observa- gard to protecting privacy will make it highly likely that they won't be studiei
tional research. But virtual worlds also have their private areas and there is For instance, if rules of informed consent are applied we'd need to:
some disagreement about the extent to which this should be taken into ac- • ascertain someone's identity (in a setting where the 'self is unstable)
count. • spend a great deal of time defining public and private (in a setting
Internet researchers can take diametrically different approaches in the where these factors are blurred)
public-private debate. A l l we can do here is to show the range of opinion. On • inform people of the nature of the research (so changing/ inhibiting it'
one hand, some people believe that making notes of online conversations is character)
eavesdropping and that this is ethically unsound however it is done. They • find ways to let people withdraw (in a setting where you can't banish then
would probably agree with Goffman that a 'conversation' includes And he asks: Should we really impose the burden of such rigid guidelines c
those most likely to increase our understanding of these new forms of life?
the right of a set of individuals once engaged in talk to have their circle
protected from entrance and overhearing by others (Goffinan 1971:64).
(b) The ownership of words debate
The offence of overhearing is, in Goffman's view of social life, a violation. It The second debate is about the 'public' or 'private' status of the words that ii
takes from an individual information not intended to be overheard, it pene- dividuals write and send to different Internet sites - basically it comes down
trates territory the speakers have claimed as their own, and it defiles that terri- whether people have a right to withhold permission for their words to be us<
tory by intruding on it. But Cavanagh (2000), who looked in depth at by others. The key question in this debate is - who owns the words that appe
Goffman's theories in relation to Internet use, used the status of "lurkers" to online?
highlight differences in online communication. To 'listen in' on online discus- Do posts belong to the poster (author), electronic group (community),
sions is at the very least tolerated - indeed many groups prefer people to listen any observer (including researchers)? (Sixsmith and Murray, 2001, p.429)
in to the group for a while before they leap in with both feet and make nai've
comments. When lurkers do 'own up' to their presence they usually receive a Is it the poster?
warm welcome and enter fully into the community. Cavanagh believes that There are two ways of looking at this - a pragmatic/legal one - and a person
such tolerance only makes sense if you see Internet interactions as occurring one.
within a public arena - you would get a different reaction if you suddenly an-
nounced your presence in someone's telephone conversation. This would be Legal
seen as eavesdropping or snooping in a 'private' conversation. One view is that all computer messages are published written material and th
to quote from this material without crediting the author is a violation of cop
Jones (1994) takes this position further. He questions whether the stress
right. However, i f we each fully owned our email messages no one could dow
on public/privacy differences in ethics guidelines make sense in a context
load or view them. In sending the message there is an 'implied licence' to rea
where rules about privacy and conditions relating to privacy are various, in-
or even archive, the information it contains. In addition, 'fair use' clauses with
consistent and not always apparent. He suggests that "forms of social life" in
copyright law allow a certain amount of material to be published if it can
cyberspace are significantly different from those in real life. They:
shown this would be to the best advantage of all concerned. So - as with so mu
• are far more ephemeral
of Internet legality - this is very vague - and returns us to the purely ethical.
• lack territorial boundaries
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

Personal not inaccurate or untrue, or likely to invoke libel laws, or against the newpa-
In the social sciences - and particularly in qualitative research and oral history per's ethical code of practice. Internet media rarely have this intermediary.
- there is an ongoing debate about the ownership of words and whether people There is sometimes a moderator but the responsibility of that person is far
will 'lose their voice' i f a researcher takes over their words. In textual research weaker than in other public media (Waern, 2001). In other words the re-
the researcher places a framework of analysis on someone else's story. The searcher is not accessing processed data - but raw data. There is no protective
danger is that people may have their views misrepresented or taken out of con- intermediary between the person who wrote the original material and the re-
text by researchers. In this situation, as Sharf asked (1999) - whose story is it searcher.
now? How has the story been changed by selecting bits for evidence or illus-
tration? Even if participants have given consent to research there is a fear they Many researchers do not see this kind of quibble as a problem. There is a
may feel they have lost control of their identity when they see their narratives widespread view that permission should not be sought for the recording and
analysis of publicly posted messages (ProjectH Research Group 1993-1994).
interpreted by others. This is made worse if people are unaware that their mes-
sages are being analyzed until the results of the research are published. People This debate is important. If material in a 'public' site belongs to an individual
who feel that individuals' own their own words, even online, think that using or a group this has implications for a researcher. Permission to use this mate-
another's words in a research study without seeking permission is an invasion rial when results are disseminated may be necessary. In the view of some re-
of privacy. They are critical of those who 'harvest' the words of unsuspecting searchers failing to reference posts and comments constitutes a form of
people for their own profit and self-aggrandizement (Sharf, 1999) plagiarism and hence exploitation and asking for permission to quote names
or some other tag is the preferred route. But this permission may be hard - if
Is it the electronic group? not impossible - to obtain in a venue where individuals may be transient - and
One longstanding newsgroup moderator talked of the 'shock' that occurred which is also used by people who may be seeking anonymity or trying out
when a writer deleted several years worth of their own posts - making new false personas. Some researchers take the robust approach that any mate-
rial appearing in a public site is freely accessible for anyone to use. Others feel
the fabric of recorded conversations, the entire history of the [group's] that just because particular online sources may be public, this does not mean
discourse...look math-eaten' (Rheingold 1993, p.36 cited in Sixsmith and
Murray 2001)
that people have waived their rights to remain anonymous, nor that the iden-
tity of an institution and/or list should be exposed.
In this view the action taken by an individual showing he 'owned' his mes- If researchers feel that it is only ethical to protect sources which have
sages destroyed the integrity of the discussion and the communal identity of provided data without prior consent - then preparing work for dissemination
the group. may be a lengthy -and ultimately a self-defeating process. First, real names,
nicknames, domain names, signatures and even ISPs may all need to be ad-
Is it the researcher? justed. Second, there may be attempts to lose demographic data which can
From a research point of view it is generally accepted that while archival ma- identify people. Third, the quotes themselves may give too much away in
terial collected from television, public records, radio, and published printed terms of speech mannerisms or contextual material and some omissions may
works might need to be referenced to the media source - it does not need the need to be made.
personal consent of the individual to use it. If the Internet is grouped with But what will happen i f the researcher sets to work to disguise and distort
these other sources then no ethical or legal standards are breached in forego- data to this degree? In some studies it may be possible to aggregate data and
ing informed consent. Many researchers believe this is the case and argue that treat it statistically (or with minimum detail) without losing its power - in
newsgroups, chat rooms and so on are as publicly accessible as a television or other studies the validity of the results will depend on the detail that has been
newspaper interview. lost. These implications in terms of reporting results of an Internet based re-
One flaw with this argument is that usually - in the public media - there's search design suggest that consideration of dissemination is something that
an intermediary between the person writing and the reader. And that person should be something that is prioritised at the beginning not the end of a
has responsible for the material presented to the public. In a newspaper article project. As this will be anti-intuitive for many researchers it is a real but nec-
the journalist will have had the responsibility of making sure comments are essary challenge for future research.

mt
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

(c) Causing harm in Interne! research debate ten in to our electronic private conversations. Reminding ourselves that emaH
As we discussed earlier 'not causing harm' is the major deontological rule that is not always destroyed when we press the delete key.
has survived the loss of old certainties. Thus the ultimate question for Internet
research must be - what harm can it do? Can researchers take the position that Legal risks
interacting online is: Lack of understanding of legal responsibities: People may see the Internet as
an uncensored, unpoliced environment but in fact researchers and participants
rather like a video game, where real world responsibilities do not apply ... no must obey many country-specific laws (such as Copyright laws and Data Pro-
matter what they say or do, it cannot harm a real human being' (Coverington, tection legislation) even online. The lack of privacy means that if these laws
1995). are broken people can find out. The lack of clarity about which legal systems
Kendall (1999) noted that the very word 'cyberspace' has a science fiction res- apply in cross-cultural research means it is difficult to know which laws will
onance - suggesting a tendency to see online worlds as independent realities apply.
separate from real world environments and histories and the living bodies that
Personal risks for participants
are susceptible to harm. In contrast, she points out that actually: 'Nobody lives
only in cyberspace' (ibid, p.70). Her view is that people move happily between Lack of protection from harassment, abuse and deception: The online world
online and offline environments (just as they move happily between different has its own forms of intimidation, harassment or threat and there are few sanc-
kinds of real life environments) and, throughout it all, they 'tend to perceive of tions and little legal redress for anyone who has negative experiences while
their identities and selves as integral and continuous' (ibid, p.60). Many re- participating in research online. In addition, playing with identity makes it
searchers acknowledge this link between the embodied thinking feeling per- technically possible to defame others using rumour, innuendo or lies or to toy
son sitting at the computer and the virtual aspects of that person interacting on with the emotions and trust of others.
the screen and feel unable to renege on their responsibility to the 'whole' per- Lack of debriefing possibilities: Some research methods may lead to anxiety
son. Seen in this way they cannot deny that Internet research may cause harm. or stress but if researchers and subjects are geographically separated from
each other and from support networks it might not be possible to give ade-
quate debriefing.
Lack of protection from exposure: Security problems increase the potential for
4 Assessment of risk in Internet research
exposure of the participant's real self, electronic self and/or their virtual com-
There seems no doubt that Internet research has risks; risks that are over and munity. This can lead to harm to reputation and self esteem - particularly
above those that researchers in conventional research have to consider. Many when sensitive topics and highly personal information are involved.
of these risks come about as a result of the technology used and although these Lack of knowledge about the potential for exposure; Newcomers to the Inter-
risks may be moderated by researchers there are still some real problems - net are not always aware of its public nature. The assumption of many partici-
such as lack of data security - which await technical solutions. There are other pants is that messages posted to an electronic forum will be seen only by a
risks which we are only just becoming aware of. Psychologically it is very small number of likeminded people. There is also a temptation for people to
possible that people 'out there' have come to harm through being involved in see discussion online as like 'a conversation on a train' and to open up in an
online research in ways that are quite unknown to us. So which risks ARE we unguarded way:
aware of?
Revelations about personal inadequacies, deviant preferences, past love affairs,
Technical risks and serious personal problems ... have passed through the [computer
Lack of security: Computer system crashes, badly set-up networks, software conferencing] system as private messages to strangers who were 'met' on the
system (Hilte and Turoff, 1993).
containing bugs, guessable passwords, viruses and hackers can all compro-
mise data Lack of protection for the most vulnerable: The difficulty of controlling the
Lack of privacy: The fact that we have no control over our electronic commu- participation of children in research not designed for them - and the techno-
nication when it's in transit - knowing that others can (and sometimes do) lis- logical and methodological difficulties in acquiring informed consent from

42 43
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

children under 18, those with learning abilities, the elderly or institutionalised. Do we recognise the autonomy of others and acknowledge that they are of
The relative anonymity of the Internet has allowed a wide variety of socially equal worth to ourselves and should be treated so?
disadvantaged and/or marginalised or stigmatised groups to express their feel- Will informed consent be sought from participants?
ings and frustrations online in ways that are emotionally difficult - or socially What procedures to obtain consent will be followed?
unwise - to do face-to-face. This has stimulated many Internet studies that How will confidentiality be respected?
have focused on all kinds of marginalised groups. There are groups with con-
cealable stigmas such as infertile women, survivors of sexual abuse, recover- These probes may seem simple and predictable. In fact, as we know, the sim-
ing drug addicts, people with non-mainstream sexualities, racists and so on. plest query in ethics may lead us into very deep water. In addition, simple
And groups with visible stigmas such as obesity, physical disability and stut- probes when related to Internet research do not produce simple answers. Once
tering. In addition, psychology researchers have flagged that the Internet is a the broad ethical issues have been defined more specific questions relating to
good place to look for individuals with certain 'disorders' or 'psychiatric con- the research could be addressed using the checklist below.
ditions' - such as a tendency to depression or panic attacks. Using people with
serious psychiatric conditions in Internet research must be considered a case
of extreme risk. 6 Checklist of ethical issues in Internet research
It is in this context of risk that active researchers need to consider the re-
sponsibilities that they hold to participants. Technological background
• Do you have a good grasp of the technology involved in the method
you will use?
J The way forward? • Do you have access to institution-based technology experts with
The Internet is a convenient and powerful research tool. It offers information whom to discuss the power and limits of the research medium?
that was previously inaccessible - or, indeed, non-existent. Taking an ethical • Will participants have the technical expertise needed for the study you
position is very challenging but we can draw on ethical guidelines that have have in mind?
been developing over years to pose some simple but systematic ethical probes • How will you organise secure storage of data, destruction of data and
(adapted with thanks from Foster, 2001) to any Internet research proposal. ownership of data?
These probes address the main ethical dilemmas in the following terms:
Organising Informed Consent
Are we seeking to magnify the good? Practical issues
Are we acting in ways that do not harm others?
• How will the consent form be accessed by participants and how will
Do we recognise the autonomy of others and acknowledge that they are
they submit it?
of equal worth to ourselves and should be treated so?
Signature on consent form
Each of these questions then lead on to more specific probes. The probes are
• Under what name (real life/ pseudonym) will people sign the consent
quite simple and clear as it is very important in ethics not to fudge the issues.
form?
• Will the consent form have a hand-written signature? I f so how will
Are we seeking to magnify the good?
you organise this?
What question is the research project addressing
• Will the signature be submitted online - how will this be done? - how
Is the research aiming at a goal which is good and desirable
will the authenticity of the signature be validated? How will the use of
What research methods will be used to achieve that goal
an online signature be defended?
How will these methods be designed to ensure the results are reliable
• What is the position with regard to the 'legally binding' nature of the
How will the results of the research be disseminated
online signature?
Are we acting in ways that do not harm others?
What is the level of risk to a participant?
How are risks assessed?

44 45
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

Description of the research • Will you repeatedly advertise your presence as a researcher - or just
• Will the aims and exact purpose(s) of the study, possible risks to par- advertise it once?
ticipants, the expected findings and the use to which such findings will • What arrangements will be made i f some members of the group won't
be put be made clear to the participants? give their consent?
• Will the researcher(s) give full information about themselves? Participant risk
• Will participants be informed fully about the methods of the research • What risks might be involved for participants in this study?
from the outset? • What measures will be taken to reduce the likelihood of risk?
I f the answer to any of the above is 'no', what kinds of information will be held • Will children and young people be involved in the study?
back or changed and why? • What arrangements will be made to protect children under 13?
• How will informed consent be arranged in anyone under 18 participates?
Clarity of description about research Withdrawal and new members
• What efforts will be made to check i f people have understood the point • What arrangements will be made to allow for withdrawal?
of the research and the research process? • How will withdrawal be handled?
• What efforts will be made to see that everyone understands the lan- • What arrangements will be made i f new people wish to enter the research?
guage of the consent form? Debriefing
• Will there be a chance for people to ask questions i f they are confused • What arrangements will be made to address participants concerns that
and/or want clarity? How?(e.g. will a FAQs facility be available to an- arise during the study?
ticipate potential questions and concerns • What arrangements will be made to debrief participants at the end of
Protection of privacy of research subjects. the study?
• What assurances about confidentiality will be made? • Will these arrangements undermine promises of anonymity/confidentiality?
• Will technology be used to secure confidentiality (e.g. passwords, en- • What facilities would be available if participants suffer i l l effects from
cryption etc)? the study?
Dissemination promises
• Are participants identifiable - is there a link between the data and the • What degree of detail will be given about the identity of the partici-
individual? pants and the venue of the research?
• IF YES - what forms of identity are involved (real names, pseudo- • Will this be made clear to participants before the study starts or will
nyms) permission be asked at the end?
• Should pseudonyms in this study be treated as real identities?

• IF NO - how will personal data be treated to make it non-identifiable. Conclusions


• [In traditional research this might involve the use of pseudonyms,
In a new research area like the Internet researchers have the responsibility of
changing any identifiable personal details (including names of places,
interpreting well established ethical guidelines used in conventional research
institutions and times), assigning code numbers, secure storage of data
in ways that are appropriate to the new online conditions. It seems clear that
and destruction of such data often after a set period of time. Online
'avoiding harm' remains an ethical constant. In addition, Internet researchers
real names, user names, domain names, signatures and even ISPs may
have additional moral obligations - to the research community. If the re-
all need to be adjusted.]
searcher is careless or lacks expertise, disillusioned and abused respondents
Working with newsgroups, online communities etc may no longer be willing to participate in research projects. Some commenta-
• How will informed consent be arranged without intruding on the inter- tors have raised the issue of 'enthusiastic well meaning DIY types' (Nancar-
action of the group? row et al, 2001) who collect data using 'the new fun toy' that is the Internet.
• Does the group have a policy about the use of members messages? These people may not know the ethical rules of professional and other bodies
• Does the group have a moderator you can get permission from? so may unwittingly break them. In some situations a junior person may be

47
Chris Mann Generating data online: ethical concerns and challenges for the C21 researcher

asked to do the research rather than a trained researcher - a person who is tech- Nancarrow, C. et al (2001) "A new research medium, new research populations and seven
nically but not ethically sawy. Some research instruments may be badly de- deadly sins for Internet researchers" pp.136-149 in ul-Haq, Rand Mann, C. (eds.)
Qualitative Market Research: An International Journal, Volume 4, Number 3, Special
signed and promises made to recipients may not be fulfilled - adding to
Issue, Research in Cyberspace.
feelings of disillusionment and alarm. Increasingly chat rooms and news-
Nosek, B. and Greenwald A. (2002) "E-Research: Ethics, security, design, and control in
groups are starting to resent the intrusion of researchers who they feel come in
Psychological research on the Internet" in Journal of Social Issues, Vol 58, No 1.
and 'strip the assets' of the group. This is bad science. For - while there are pp.I61-176.
many different ways to do research - a research method which leaves behind Rheingold, H (1993) The Virtual Community: Homesteading on the electric frontier. Reading,
bad feeling is morally suspect in terms of both participants and other research- MA: Addison-Wesley.
ers. Researchers who are guilty of negligence are acting in ways that will Seidman, I. (1991). Interviewing as Qualitative Research. Columbia University: Teacher's
harm others - not just the individuals they are working with - but also because College.
they close down opportunities for valuable research to be conducted in the fu- Senior, C. and Smith, M. (1999) "The Internet... a possible research tool?" in Psychologist,
ture. 12, pp.442-44.
Sharf, B. (1999) "Beyond N etiquette, The ethics of doing naturalistic discourse research on the
Internet" pp.243-256 in Jones, S Doing Internet Research, Sage, Thousand Oaks
Sixsmith J. and Murray C. (2001) "Ethical issues in the documentary data analysis of Internet
Bibliography posts and archives" in Qualitative Health Research, Vol.11, No.3, May
Smith, M. and Leigh, B. (1997) "Virtual subjects: using the Internet as an alternative source of
Adler, P.A. and Adler, P. (2001) "The reluctant respondent" in (eds.) Gubrium, J. And subjects and research environment." Behaviour, Research Methods, Instruments, and
Computers, 1997, 29(4), 496-505.
Holstein, I. Handbook of interview Research, pp.515-535, London, Thousand
Van Gelder, L. (1991). "The strange case of the electronic lover." in C. Dunlop and R. filing
Oaks:Sage (eds.), Computerization and Controversy: Value Conflicts and Social Choices. New
Bimbaum, M. (2000) (ed.) Psychological Experiments on the Internet, Academic press, San York: Academic Press.
Diego
Cavanagh, A. (2000) "Behaviour in public?: ethics in online ethnography" in Research
Methodology Online, Issue 6. CHRIS MANN

Coverington, M (1995) "Design and implementation of a campus computer ethics policy" in Faculty of Social and Political Sciences, University of Cambridge, England
ccm 10@cam.ac.uk
Internet research Electronic Networking Applications and Policy. Vol.5. No 4, pp.31-
41.
Foster, C. (2001) The Ethics of Medical Research on Humans. Cambridge University Press
Goffman, E (1971) Relations in Public: Microstudies of the Public Order, Penguin:
Harmondsworth.
Hiltz, S. and TurofF, M. (1993). The Network Nation: Human communication via computer.
Reading, MA: MIT Press.
http://www.emerald-library.com/brev/17204ccl.htm
Jones, A. (1994) "The Ethics of Research in Cyberspace" in Internet Research, Vol. 4, No. 3,
1994, pp. 30-35.
Kendall, L. (1999). "Recontextualizing cyberspace: Methodological considerations for on-line
research", in (ed.) S. Jones, Doing Internet Research pp. 57-75. Thousand Oaks, CA,
and London: Sage.
Mann, C and Stewart, F. (2000) Using the Internet in Qualitative Research: A Handbook for
Researching Online. New Technologies for Social Research series, pp. 39-62.
London: Sage.
(2001) 'Internet Interviewing' in Gubrium, J. and Holstein, J. (eds.) Handbook of
Interview Research. London, Thousand Oaks: Sage.

48 49

Das könnte Ihnen auch gefallen